A method proposed by Clements and Sjaastad for measuring the effect of manufacturing protection on farmers' incomes is shown to be theoretically unsound, to use some inappropriate assumptions and to lead to overestimation. Even if corrected, the method seems inferior to estimating directly via ORANI the change in farm incomes which would occur when protection levels change. This is because the model contains detailed theoretical explanations of the price and quantity changes which determine changes in the relevant incomes and is calibrated using extensive empirical evidence about the behaviour of the prices and quantities in the Australian economy. Using ORANI and incorporating recently completed improvements to its agricultural data base, it is estimated that, in the short run, abolishing protection would increase farm incomes by about 17 per cent.